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 .
This office action is in response to the claimed amendment filed on November 07, 2025, in which claims 1-20 are presented for examination.
Information Disclosure Statement
The information disclosure statement filed on November 07, 2025, complies with the provisions of 37 CFR 1.97, 1.98 and MPEP § 609. It has been placed in the application file. The information referred to therein has been considered as to the merits.
Claim Rejections - 35 USC § 101
35 U.S.C. 101 reads as follows:
Whoever invents or discovers any new and useful process, machine, manufacture, or composition of matter, or any new and useful improvement thereof, may obtain a patent therefor, subject to the conditions and requirements of this title.
Claims 1-20 are rejected under 35 U.S.C. 101 because the claimed invention is directed to an abstract without significantly more.
Step 1, Statutory Category:
Claims 1-13 are directed to a method
Claims 14-19 are directed to a computer.
Claim 20 is directed to a system.
Therefore, claims 1-20 fall into at least one of the four statutory categories of invention, i.e., process, machine, manufacture, or composition of matter.
Step 2A, Prong One (Judicial exception recited)
The limitation “determining, by the computer, a segmentation result based on the query using a first language model” in claims 1, 14 and 20, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. One can mentally and/or manually with the aid of pen and paper determine a segmentation result.
The limitation “determining, by the computer, a query embedding based on the query” in claims 1, 14 and 20, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. One can mentally and/or manually with the aid of pen and paper determine a query embedding.
The limitation “determining, by the computer, a plurality of candidate labels based on the query embedding” in claims 1, 14 and 20, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. One can mentally and/or manually with the aid of pen and paper determine a plurality of candidate labels.
The limitation “determining, by the computer, a linking result based on the query and the plurality of candidate labels using a second language model” in claims 1, 14 and 20, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. One can mentally and/or manually with the aid of pen and paper determine a linking result.
The limitation “providing, by the computer, the one or more documents to the end user device” in claims 1, 14 and 20, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement, but for the recitation of generic computer components. One can mentally and/or manually with the aid of pen and paper provide the one or more documents to the end user device.
Step 2A, Prong Two (Integrated into a practical application):
This judicial exception is not integrated into a practical application. In particular, the claims recite the following additional elements:
The limitation “receiving, by a computer, a query from an end user device” amounts to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)).
The limitation “retrieving, by the computer, one or more documents from a documents database using the linking result and the segmentation result” amounts to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)).
The limitation “a processor, an end user device; a documents database; a central server computer; and a non-transitory computer readable medium” are recited at a high level of generality such that they amount to on more than mere instructions to apply the exception using a generic component. (see MPEP 2106.05(f)). These limitations can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of a computer (see MPEP 2106.05(h)). Note, the mere instructions to apply an exception on a generic computer cannot integrate a judicial exception into a practical application.
Step 2B (claim provides an inventive concept):
With respect to the “receiving …; and retrieving ….” identified as insignificant extra-solution activity above when re-evaluated this element is well-understood, routine, and conventional as evidenced by the court cases in MPEP 2106.05(d)(II), "i. Receiving or transmitting data over a network, e.g., using the Internet to gather data, Symantec, 838 F.3d at 1321, 120 USPQ2d at 1362 (utilizing an intermediary computer to forward information); … OIP Techs., Inc., v. Amazon.com, Inc., 788 F.3d 1359, 1363, 115 USPQ2d 1090, 1093 (Fed. Cir. 2015) (sending messages over a network); buySAFE, Inc. v. Google, Inc., 765 F.3d 1350, 1355, 112 USPQ2d 1093, 1096 (Fed. Cir. 2014) (computer receives and sends information over a network);" and thus remains insignificant extra-solution activity that does not provide significantly more.
With respect to the “a processor, an end user device; a documents database; a central server computer; and a non-transitory computer readable medium” amount to elements that have been recognized as well-understood, routine, and conventional activity in particular fields, as demonstrate by: Relevant court decision: the followings are examples of court decisions demonstrating well-understood, routine and conventional activities, see e.g., MPEP 2106.05(d)(II) and MPEP 2106.05(f)(2): Computer readable storage media comprising instructions to implement a method, e.g., see Versata Dev. Group, Inc. v. SAP Am., Inc., 793 F.3d 1306, 1334, 115 USPQ2d 1681, 1701 (Fed. Cir. 2015).
The claims do not include additional elements that are sufficient to amount to significantly more than the judicial exception because the additional elements when considered both individually and as an ordered combination do not amount to significantly more than the abstract idea.
Looking at the claim, as a whole, does not change this conclusion and the claim appears to be ineligible.
Accordingly, claim 1 is directed to an abstract idea. The remaining independent claims 14 and 20 fall short the 35 USC 101 requirement under the same rationale.
The dependent claims 2-12 and 15-19 when analyzed and each taken as a whole are held to be patent ineligible under 35 USC 101 because the additional recited limitations fail to establish that the claims are not directed to an abstract idea.
Claim 2 recites “the coordinated storage management functionality comprises applying coordinated file system operations to the contents of the first and second managed directories as a group”. This additional element is recited at a high level of generality and would function in its ordinary capacity for applying coordinated file system operations, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 2 recites “determining, by the computer, the query embedding using an embedding machine learning model that is trained to generate embeddings”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 3 recites “providing, by the computer, the fulfillment request message to a service provider computer operated by the service provider, wherein the service provider initiates preparation of at least the item; determining, by the computer, one or more transporter user devices; providing, by the computer, the fulfillment request message to the one or more transporter user devices, wherein the one or more transporter user devices determine whether or not to request to accept the fulfillment request message; generating, by the computer, an update message indicating a status of the fulfillment request message; and providing, by the computer, the update message to the end user device”. These limitations, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception. The limitations “receiving, by the computer from the end user device, a fulfillment request message comprising at least one item provided by a service provider associated with the image; and receiving, by the computer, an acceptance message from a transporter user device of the one or more transporter user devices” amount to data-gathering steps which is considered to be insignificant extra-solution activity, (See MPEP 2106.05(g)).
Claim 4 recites “determining, by the computer, a category template for the query; generating, by the computer, a prompt using the query and the category template; and providing, by the computer, the prompt to the first language model to generate the segmentation result in a particular format based on the query”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 5 recites “determining, by the computer, a taxonomy category for the query; and determining, by the computer, the category template based on the taxonomy category”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 6 recites “comparing, by the computer, the query embedding to a plurality of labels stored in a label database to determine a plurality of similarity scores; and selecting, by the computer, a subset of labels from the plurality of labels based on the plurality of similarity scores, wherein the subset of labels is the plurality of candidate labels”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 7 recites “generating, by the computer, a prompt based on the query and the plurality of candidate labels; and determining, by the computer, the linking result using the second language model and the prompt.”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 8 recites “wherein the linking result includes one or more known labels that represent the query, wherein the one or more known labels are labels that are indexed in the document database, and wherein the segmentation result includes one or more generated labels”. This additional element is recited at a high level of generality and would function in its ordinary capacity for linking result, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 9 recites “obtaining, by the computer, retrieved documents from the documents database using the one or more known labels that represent the query of the linking result; and filtering, by the computer, the retrieved documents using the segmentation result to determine the one or more documents”. This additional element is recited at a high level of generality and would function in its ordinary capacity for obtaining retrieved documents from the documents database, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 10 recites “for each document of the retrieved documents, determining, by the computer, whether or not the document includes text that matches the one or more generated labels in the segmentation result; and if the document includes text that matches the one or more generated labels in the segmentation result, adding the document to the one or more documents that are to be provided to the end user device”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 11 recites “ranking, by the computer, the one or more documents using the segmentation result”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 12 recites “after receiving the query, preprocessing, by the computer, the query”. This additional element is recited at a high level of generality and would function in its ordinary capacity for preprocessing, by the computer, the query, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 13 recites “modifying, by the computer, the query to fix spelling of words in the query”. This additional element is recited at a high level of generality and would function in its ordinary capacity for modifying the query to fix spelling of words in the query, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more. The limitation “determining, by the computer, a search intent of the query using a search intent classification process”, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception
Claim 15 recites “wherein the query includes a search for an item, wherein the search includes text and wherein the one or more documents represent items”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 16 recites “wherein the plurality of candidate labels include cuisine related labels, dish type related labels, and/or dietary preference related labels”. This limitation, as drafted, is a process that, under its broadest reasonable interpretation, covers a mental process as a form of evaluation or judgement. There is no additional elements recited which tie the abstract idea into a practical application and does not amount to significant more than the identified judicial exception.
Claim 17 recites “wherein receiving the query comprises: receiving the query from an application installed on the end user device, and wherein providing the one or more documents to the end user device comprises: providing the one or more documents to the application”. This additional element is recited at a high level of generality and would function in its ordinary capacity for receiving the query from an application installed on the end user device, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 18 recites “wherein the segmentation result includes one or more generated labels, and wherein determining the segmentation result comprises: determining a taxonomy category for the query; determining a category template based on the taxonomy category; generating a prompt using the query and the category template; and providing the prompt to the first language model to generate the segmentation result”. This additional element is recited at a high level of generality and would function in its ordinary capacity for receiving the query from an application installed on the end user device, this additional element does not integrate the integrate the judicial exception into a practical application and does not amount to significantly more.
Claim 19 recites “wherein the computer is a central server computer” are recited at a high level of generality such that they amount to on more than mere instructions to apply the exception using a generic component. (see MPEP 2106.05(f)). These limitations can also be viewed as nothing more than an attempt to generally link the use of the judicial exception to the technological environment of a computer (see MPEP 2106.05(h)). Note, the mere instructions to apply an exception on a generic computer cannot integrate a judicial exception into a practical application.
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 1-20 are rejected under 35 U.S.C. 103 as being unpatentable over He et al., (hereinafter “He”) US 20200218722 A1 in view of Ahmadvand et al., (hereinafter “Ahmadvand”) US 20210110208 A1.
As claim 1 He discloses a method comprising:
receiving, by a computer, a query from an end user device (see [0003], receiving a query, wherein the query includes a first sequence of words);
determining, by the computer, a segmentation result based on the query using a first language model (see [0003], converting the query into a second sequence of words by using a first machine learning model);
determining, by the computer, a query embedding based on the query (see [0032], the system 102 may feed questions 305 or queries to the first algorithm group 320 and the second algorithm group 330 to obtain answers 311 or rank of sentences 317 as results 244 to the questions 305 or queries. The questions 305 or queries are natural language, such as “how are you today?” “what are their most popular drinks?” “are there any good vegan choices?” etc. The system 102 may first feed the questions 305 to a programmer 322. A programmer 322 may be a machine learning model, which is a set of code or instructions, trained based on training data and executable by one or more processors to perform predetermined functions. For example, the programmer 322 may be a first sequence to sequence machine learning model 322 executable by one or more processors to covert a sequence to another sequence. A sequence may be a series of numbers or characters. In some embodiments, a programmer 322 may be in hardware form or in a mixed form of hardware and software. The programmer 322 may also be configured to perform the functions described below. The first sequence to sequence machine learning model 322, such as the programmer 322, may convert one sequence of words to another sequence of words with the same or a different length (i.e., the number of words). By deriving the relations from the previous words in the same sequence, the sequence to sequence model may be trained to pick a current word with the highest probability from a large pool of words. The sequence to sequence machine learning model 322 may be one of a Long short-term Memory (LSTM) network, a Recurrent Neural network (RNN), a Gated Recurrent Unit (GRU) network, etc. In some embodiments, The programmer 322 may be trained to convert a question 305 (e.g., a sequence of words) to an n-gram 307 such as a bigram (e.g., a contiguous sequence of two words). In the above example where the question 305 is “are there any good vegan choices?”, the programmer 322 may be trained to convert the question 305 to a sequence of words, e.g., “tofu wings.” Each of the words, e.g., “tofu,” “wings,” may correspond to the highest probability among a large word pool, and the word “wings” may be determined upon the determination of “tofu,” which means the relationship between previous word “tofu” and the latter word “wings” may be factored into determining the latter word “wings.”);
determining, by the computer, a linking result based on the query and the plurality of candidate labels using a second language model (see [0008]-[0009], obtaining a result for the query by applying a second machine learning model to a combination of the first sequence of words and the second sequence of words);
retrieving, by the computer, one or more documents from a documents database using the linking result and the segmentation result (see [0005], obtaining a result for the query by applying a second sequence to sequence model to a combination of the first sequence of words and the second sequence of words comprises: feeding the third sequence of words into the second machine learning model to obtain a fourth sequence of words; and generating the result for the query based on the fourth sequence of words, wherein the first and second machine learning models are trained based on training data comprising: a plurality of queries, a plurality of sentences, and a plurality of results, and wherein the plurality of sentences are retrieved from unstructured data); and
providing, by the computer, the one or more documents to the end user device (see [0005]-[0009).
Ahmadvand discloses the claimed “determining, by the computer, a plurality of candidate labels based on the query embedding” (see [0025]-[0032], initializing with popular word embeddings associated with the search engine on which the trained model will be applied; and concatenates the first set of candidate labels and second set of candidate labels to generate a concatenated set of labels. Concatenating the sets of candidate labels may include concatenating a matrix of the first candidate labels with a matrix of the second candidate labels, in an embodiment, to generate a vector representative of the entire candidate label space).
Therefore, it would have been obvious to one having ordinary in the art before the effective filing date of the claimed invention to combine the system of He and Ahmadvand to arrive the invention as claimed in order to ensure simple and efficient segmentation process, thereby increasingly formulating queries that combine multiple attributes and constraints by users and reducing computational cost while maintaining high accuracy.
As to claim 2, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed “determining the query embedding comprises: determining, by the computer, the query embedding using an embedding machine learning model that is trained to generate embeddings” (see [0040]-[0045]).
As to claim 3, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed “receiving, by the computer from the end user device, a fulfillment request message comprising at least one item provided by a service provider associated with the image; providing, by the computer, the fulfillment request message to a service provider computer operated by the service provider, wherein the service provider initiates preparation of at least the item; determining, by the computer, one or more transporter user devices; providing, by the computer, the fulfillment request message to the one or more transporter user devices, wherein the one or more transporter user devices determine whether or not to request to accept the fulfillment request message; receiving, by the computer, an acceptance message from a transporter user device of the one or more transporter user devices; generating, by the computer, an update message indicating a status of the fulfillment request message; and providing, by the computer, the update message to the end user device”.
As to claim 4, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed “determining, by the computer, a category template for the query; generating, by the computer, a prompt using the query and the category template; and providing, by the computer, the prompt to the first language model to generate the segmentation result in a particular format based on the query” (see [0040]-[0045]).
As to claim 5, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed “determining, by the computer, a taxonomy category for the query; and determining, by the computer, the category template based on the taxonomy category” (see [0040]-[0045]).
As to claim 6, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed wherein determining the plurality of candidate labels comprises: comparing, by the computer, the query embedding to a plurality of labels stored in a label database to determine a plurality of similarity scores; and selecting, by the computer, a subset of labels from the plurality of labels based on the plurality of similarity scores, wherein the subset of labels is the plurality of candidate labels” (see par. [0045]-[0048]).
As to claim 7, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed wherein determining the linking result comprises: generating, by the computer, a prompt based on the query and the plurality of candidate labels; and determining, by the computer, the linking result using the second language model and the prompt (see [0041] In FIG. 4A, graph 402 represents a sentence (e.g., the sentence 301). As shown in FIG. 4B, a first sentence 402 may be “After scanning the menu for a bit however, I was able to find the tofu wings.” A second sentence 402 may be “On my first visit, I was there as a vegan, yeah you read that right.” In FIG. 4A, circle 406 represents an n-gram space or knowledge graph (e.g., the n-gram space 303) obtained from multiple sentences 402. In the n-gram space 406, there may be a number of n-grams, which are graphically represented by blocks 404. These n-grams 404 have been generated from the one or more sentences 402. Graph 408 represents a question (e.g., the question 305) inputted by a user. As shown in FIG. 4B, for example, the question 408 may be “Any good vegan choices?”
[0042] Referring back to FIG. 4A, through a sequence to sequence model (e.g., the first sequence to sequence model 322), an n-gram 410 may be obtained given the question 408. For example, the n-gram 410 may correspond to the highest log-likelihood. The first sequence to sequence model 322 may also check whether the n-gram 410 is within the n-gram space 406. As shown in FIG. 4B, the n-gram 410 may be a bigram, e.g., “Tofu wings.” This bigram 410 “Tofu wings” may be obtained from the first sentence 402—“After scanning the menu for a bit however, I was able to find the tofu wings,” and correspond to the highest log-likelihood. In the embodiments shown in FIG. 4B, the second exemplary sentence—“On my first visit, I was there as a vegan, yeah you read that right”—may also be used to generate multiple bigrams. However, none of the bigrams generated from the second sentence correspond to the highest log-likelihood based on the first sequence to sequence model 322, and thus may not be selected by the model 322.
[0043] In some embodiments, based on the question 408 and the n-gram 410, an answer 412 may be generated based on a second sequence to sequence model 334 as described with reference to FIG. 3. As shown in FIG. 4B, the answer 412 may be “Tofu wings could be a choice.” In contrast, since no n-gram has been chosen from the second sentence, the second sentence then may not be able to yield an answer. Referring back to FIG. 4A, a score 416 may also be obtained based on the question 408 and the n-gram 410, indicating the relevance between the sentence 402 and the question 408. For example, as shown in FIG. 4B, the pair of the question 408 and the first sentence may have a score of “0.8,” indicating the first sentence answers the question 408 well, while the pair of the question 408 and the second sentence may have a score of “0.1,” indicating the second sentence may not be able to answer the question 408. Referring back to FIG. 4A, the scores 416 associated with sentences 402 may be used as one of the criteria to rank the sentences 402. As shown in FIG. 4B, the first sentence may have a higher rank than the second sentence.
[0044] FIG. 5 illustrates a flowchart of an exemplary method for query responding, in accordance with various embodiments. The method 500 may be implemented in various environments including, for example, the environment 100 of FIG. 1, or the exemplary system 200 of FIG. 2. The exemplary method 500 may be implemented by one or more components of the system 102 (e.g., the processor 104, the memory 106). The exemplary method 500 may be implemented by multiple systems similar to the system 102. The operations of method 500 presented below are intended to be illustrative. Depending on the implementation, the exemplary method 500 may include additional, fewer, or alternative steps performed in various orders or in parallel).
As to claim 8, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed wherein the linking result includes one or more known labels that represent the query, wherein the one or more known labels are labels that are indexed in the document database, and wherein the segmentation result includes one or more generated labels (see [0041] In FIG. 4A, graph 402 represents a sentence (e.g., the sentence 301). As shown in FIG. 4B, a first sentence 402 may be “After scanning the menu for a bit however, I was able to find the tofu wings.” A second sentence 402 may be “On my first visit, I was there as a vegan, yeah you read that right.” In FIG. 4A, circle 406 represents an n-gram space or knowledge graph (e.g., the n-gram space 303) obtained from multiple sentences 402. In the n-gram space 406, there may be a number of n-grams, which are graphically represented by blocks 404. These n-grams 404 have been generated from the one or more sentences 402. Graph 408 represents a question (e.g., the question 305) inputted by a user. As shown in FIG. 4B, for example, the question 408 may be “Any good vegan choices?” [0042] Referring back to FIG. 4A, through a sequence to sequence model (e.g., the first sequence to sequence model 322), an n-gram 410 may be obtained given the question 408. For example, the n-gram 410 may correspond to the highest log-likelihood. The first sequence to sequence model 322 may also check whether the n-gram 410 is within the n-gram space 406. As shown in FIG. 4B, the n-gram 410 may be a bigram, e.g., “Tofu wings.” This bigram 410 “Tofu wings” may be obtained from the first sentence 402—“After scanning the menu for a bit however, I was able to find the tofu wings,” and correspond to the highest log-likelihood. In the embodiments shown in FIG. 4B, the second exemplary sentence—“On my first visit, I was there as a vegan, yeah you read that right”—may also be used to generate multiple bigrams. However, none of the bigrams generated from the second sentence correspond to the highest log-likelihood based on the first sequence to sequence model 322, and thus may not be selected by the model 322.
[0043] In some embodiments, based on the question 408 and the n-gram 410, an answer 412 may be generated based on a second sequence to sequence model 334 as described with reference to FIG. 3. As shown in FIG. 4B, the answer 412 may be “Tofu wings could be a choice.” In contrast, since no n-gram has been chosen from the second sentence, the second sentence then may not be able to yield an answer. Referring back to FIG. 4A, a score 416 may also be obtained based on the question 408 and the n-gram 410, indicating the relevance between the sentence 402 and the question 408. For example, as shown in FIG. 4B, the pair of the question 408 and the first sentence may have a score of “0.8,” indicating the first sentence answers the question 408 well, while the pair of the question 408 and the second sentence may have a score of “0.1,” indicating the second sentence may not be able to answer the question 408. Referring back to FIG. 4A, the scores 416 associated with sentences 402 may be used as one of the criteria to rank the sentences 402. As shown in FIG. 4B, the first sentence may have a higher rank than the second sentence. [0044] FIG. 5 illustrates a flowchart of an exemplary method for query responding, in accordance with various embodiments. The method 500 may be implemented in various environments including, for example, the environment 100 of FIG. 1, or the exemplary system 200 of FIG. 2. The exemplary method 500 may be implemented by one or more components of the system 102 (e.g., the processor 104, the memory 106). The exemplary method 500 may be implemented by multiple systems similar to the system 102. The operations of method 500 presented below are intended to be illustrative. Depending on the implementation, the exemplary method 500 may include additional, fewer, or alternative steps performed in various orders or in parallel).
As to claim 9, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed wherein retrieving the one or more documents from the documents database using the linking result and the segmentation result comprises:
obtaining, by the computer, retrieved documents from the documents database using the one or more known labels that represent the query of the linking result; and filtering, by the computer, the retrieved documents using the segmentation result to determine the one or more documents (see [0035] In some embodiments, the first algorithm group 320 may determine 324 whether the n-gram 307 outputted from the programmer 322 is within the n-gram space 303 of the sentences 301. If the n-gram 307 is out of the n-gram space 303 of the sentences 301, a meaningful answer or result may not be obtained for the question 305. If the first algorithm group 320 determines 324 that the n-gram 307 is within the n-gram space 303 of the sentences 301, then the first algorithm group 320 may output the n-gram 307 to the second algorithm group 330 for further processing.
[0036] The second algorithm group 330 may be configured to receive the n-gram 307 and combine the question 305 and the n-gram 307. In some embodiments, the second algorithm group 330 may concatenate the question 305 with the n-gram 307 to obtain a concatenated sequence 309. For example, if the question 305 is “any good vegan choice?” and the bigram 307 outputted by the programmer 322 is “tofu wings,” then the concatenated sequence 309 may be “any good vegan choice? tofu wings.” In a computer language, the concatenated sequence 309 may be represented by [“any”, “good”, “vegan”, “choices”, “?”, “<QSSEP>”, “tofu”, “wings”]. The second algorithm group 330 may feed the concatenated sequence 309 into an answerer 334. Similar to the programmer 322, an answerer 334 may also be a machine learning model, which is a set of code or instructions, trained based on training data and executable by one or more processors to perform predetermined functions. For example, the answerer 334 may be a second sequence to sequence model. In some embodiments, an answerer 334 may be in hardware form or in a mixed form of hardware and software. The answerer 334 may also be configured to perform the functions described below. For example, the answerer 334 may be trained to generate an answer to the question 305 if the n-gram 307 can answer the question 305 (which may also means that one or more sentences 301 tied to the n-gram 307 such as those sentences 301 including the n-gram 307 can be related to or answer the question 305). The answerer 334 may return no answer to the question 305 if the n-gram 307 cannot answer the question 305 (which means that no collected sentence 301 is able to answer the question 305). In the above example where the question 305 is “any good vegan choice?”, if the bigram 307 is “scan menu,” then the concatenated sequence 309 may be “any good vegan choice? scan menu).
As to claim 10, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed wherein filtering the retrieved documents comprises:
for each document of the retrieved documents, determining, by the computer, whether or not the document includes text that matches the one or more generated labels in the segmentation result; and if the document includes text that matches the one or more generated labels in the segmentation result, adding the document to the one or more documents that are to be provided to the end user device (see [0035] In some embodiments, the first algorithm group 320 may determine 324 whether the n-gram 307 outputted from the programmer 322 is within the n-gram space 303 of the sentences 301. If the n-gram 307 is out of the n-gram space 303 of the sentences 301, a meaningful answer or result may not be obtained for the question 305. If the first algorithm group 320 determines 324 that the n-gram 307 is within the n-gram space 303 of the sentences 301, then the first algorithm group 320 may output the n-gram 307 to the second algorithm group 330 for further processing.
[0036] The second algorithm group 330 may be configured to receive the n-gram 307 and combine the question 305 and the n-gram 307. In some embodiments, the second algorithm group 330 may concatenate the question 305 with the n-gram 307 to obtain a concatenated sequence 309. For example, if the question 305 is “any good vegan choice?” and the bigram 307 outputted by the programmer 322 is “tofu wings,” then the concatenated sequence 309 may be “any good vegan choice? tofu wings.” In a computer language, the concatenated sequence 309 may be represented by [“any”, “good”, “vegan”, “choices”, “?”, “<QSSEP>”, “tofu”, “wings”]. The second algorithm group 330 may feed the concatenated sequence 309 into an answerer 334. Similar to the programmer 322, an answerer 334 may also be a machine learning model, which is a set of code or instructions, trained based on training data and executable by one or more processors to perform predetermined functions. For example, the answerer 334 may be a second sequence to sequence model. In some embodiments, an answerer 334 may be in hardware form or in a mixed form of hardware and software. The answerer 334 may also be configured to perform the functions described below. For example, the answerer 334 may be trained to generate an answer to the question 305 if the n-gram 307 can answer the question 305 (which may also means that one or more sentences 301 tied to the n-gram 307 such as those sentences 301 including the n-gram 307 can be related to or answer the question 305). The answerer 334 may return no answer to the question 305 if the n-gram 307 cannot answer the question 305 (which means that no collected sentence 301 is able to answer the question 305). In the above example where the question 305 is “any good vegan choice?”, if the bigram 307 is “scan menu,” then the concatenated sequence 309 may be “any good vegan choice? scan menu).
As to claim 11, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed ranking, by the computer, the one or more documents using the segmentation result (see [0031] In some embodiments, the system 102 may feed sentences 301 (e.g., sentences from POI reviews) to the search and answering engine 226 to generate 310 multiple n-grams. An n-gram is a contiguous sequence of “n” items from a given sample of text or speech, where “n” can be any positive integers. For example, an n-gram generated 310 by the search and answering engine 226 may be a contiguous sequence of “n” words in a sentence 301. In some embodiments, the search and answering engine 226 may generate 310 all possible n-grams from each sentence 301 to form an n-gram space 303. For example, the search and answering engine 226 may generate 310 all possible bigrams (e.g., contiguous sequences of two words) from each sentence 301 to form a bigram space 303. With respect to a sentence 301 (e.g., “we were there for about two hours”), the search and answering engine 226 may generate 310 bigrams such as “we were,” “were there,” “there for,” “for about,” “about two,” “two hours,” etc. The bigram space 303 may include all the bigrams generated from the sentence 301. Other types of n-grams may be generated and used to obtain the n-gram space 303, e.g., unigram, trigram, etc.
[0032] In some embodiments, the system 102 may feed questions 305 or queries to the first algorithm group 320 and the second algorithm group 330 to obtain answers 311 or rank of sentences 317 as results 244 to the questions 305 or queries. The questions 305 or queries are natural language, such as “how are you today?” “what are their most popular drinks?” “are there any good vegan choices?” etc. The system 102 may first feed the questions 305 to a programmer 322. A programmer 322 may be a machine learning model, which is a set of code or instructions, trained based on training data and executable by one or more processors to perform predetermined functions. For example, the programmer 322 may be a first sequence to sequence machine learning model 322 executable by one or more processors to covert a sequence to another sequence. A sequence may be a series of numbers or characters. In some embodiments, a programmer 322 may be in hardware form or in a mixed form of hardware and software. The programmer 322 may also be configured to perform the functions described below. The first sequence to sequence machine learning model 322, such as the programmer 322, may convert one sequence of words to another sequence of words with the same or a different length (i.e., the number of words). By deriving the relations from the previous words in the same sequence, the sequence to sequence model may be trained to pick a current word with the highest probability from a large pool of words. The sequence to sequence machine learning model 322 may be one of a Long short-term Memory (LSTM) network, a Recurrent Neural network (RNN), a Gated Recurrent Unit (GRU) network, etc. In some embodiments, The programmer 322 may be trained to convert a question 305 (e.g., a sequence of words) to an n-gram 307 such as a bigram (e.g., a contiguous sequence of two words). In the above example where the question 305 is “are there any good vegan choices?”, the programmer 322 may be trained to convert the question 305 to a sequence of words, e.g., “tofu wings.” Each of the words, e.g., “tofu,” “wings,” may correspond to the highest probability among a large word pool, and the word “wings” may be determined upon the determination of “tofu,” which means the relationship between previous word “tofu” and the latter word “wings” may be factored into determining the latter word “wings.”).
As to claim 12, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed after receiving the query, preprocessing, by the computer, the query.
As to claim 13, the combination of He and Ahmadvand discloses the invention as claimed. In addition, He discloses the claimed, wherein preprocessing the query comprises: modifying, by the computer, the query to fix spelling of words in the query; and determining, by the computer, a search intent of the query using a search intent classification process (see [0031] In some embodiments, the system 102 may feed sentences 301 (e.g., sentences from POI reviews) to the search and answering engine 226 to generate 310 multiple n-grams. An n-gram is a contiguous sequence of “n” items from a given sample of text or speech, where “n” can be any positive integers. For example, an n-gram generated 310 by the search and answering engine 226 may be a contiguous sequence of “n” words in a sentence 301. In some embodiments, the search and answering engine 226 may generate 310 all possible n-grams from each sentence 301 to form an n-gram space 303. For example, the search and answering engine 226 may generate 310 all possible bigrams (e.g., contiguous sequences of two words) from each sentence 301 to form a bigram space 303. With respect to a sentence 301 (e.g., “we were there for about two hours”), the search and answering engine 226 may generate 310 bigrams such as “we were,” “were there,” “there for,” “for about,” “about two,” “two hours,” etc. The bigram space 303 may include all the bigrams generated from the sentence 301. Other types of n-grams may be generated and used to obtain the n-gram space 303, e.g., unigram, trigram, etc.
[0032] In some embodiments, the system 102 may feed questions 305 or queries to the first algorithm group 320 and the second algorithm group 330 to obtain answers 311 or rank of sentences 317 as results 244 to the questions 305 or queries. The questions 305 or queries are natural language, such as “how are you today?” “what are their most popular drinks?” “are there any good vegan choices?” etc. The system 102 may first feed the questions 305 to a programmer 322. A programmer 322 may be a machine learning model, which is a set of code or instructions, trained based on training data and executable by one or more processors to perform predetermined functions. For example, the programmer 322 may be a first sequence to sequence machine learning model 322 executable by one or more processors to covert a sequence to another sequence. A sequence may be a series of numbers or characters. In some embodiments, a programmer 322 may be in hardware form or in a mixed form of hardware and software. The programmer 322 may also be configured to perform the functions described below. The first sequence to sequence machine learning model 322, such as the programmer 322, may convert one sequence of words to another sequence of words with the same or a different length (i.e., the number of words). By deriving the relations from the previous words in the same sequence, the sequence to sequence model may be trained to pick a current word with the highest probability from a large pool of words. The sequence to sequence machine learning model 322 may be one of a Long short-term Memory (LSTM) network, a Recurrent Neural network (RNN), a Gated Recurrent Unit (GRU) network, etc. In some embodiments, The programmer 322 may be trained to convert a question 305 (e.g., a sequence of words) to an n-gram 307 such as a bigram (e.g., a contiguous sequence of two words). In the above example where the question 305 is “are there any good vegan choices?”, the programmer 322 may be trained to convert the question 305 to a sequence of words, e.g., “tofu wings.” Each of the words, e.g., “tofu,” “wings,” may correspond to the highest probability among a large word pool, and the word “wings” may be determined upon the determination of “tofu,” which means the relationship between previous word “tofu” and the latter word “wings” may be factored into determining the latter word “wings.”).
As to claims 14-19, claims 14-19 are computer for performing the method of claims 1-13. They are rejected under the same rationale.
As to claim 20, claim 20 is a system for performing the method of claims 1-13. It is rejected under the same rationale.
Conclusion
The prior art made of record and not relied upon is considered pertinent to applicant's disclosure.
US 10467673 B2 (involved in determining that the query includes a first query term associated with a first product, wherein the first product is identified by a first product code, substitutes the first product code for the first query term in a first intermediate query, and searches the product information database based upon the first product code.)
US 20170249683 (involved in receiving a query at a query handling system. A determination is made that the query includes a first query term associated with a first product, where the first product is identified by a first product code. The first product code for the first query term is substituted in a first intermediate query. A product information database is searched based upon the first product code. A search result is provided in response to searching the product information database, where the search result includes the first product.)
US 20150193510 A1 (involved in detecting a user activation of a document operation search region. A set of frequently used document operations is identified using a database stored on a server in response to detecting the activation. A set of document operation labels corresponding to the identified document operations is displayed. User selection of a respective document operation label of the set of document operation labels is detected. The respective document operation corresponding to the selected respective document operation label is performed.)
US 20130226935 A1 (involved in facilitating context-based search query formation.)
US 20120084321 (involved in receiving a user-specified text string that includes multiple search keywords, the computer identifies a chunk within the document that satisfies the search keywords and displays the identified chunk to the user, wherein terms in the identified chunk that match the search keywords are either ordered differently from the search keywords in the user-specified text string or separated from one another by at least one term not matching any of the search keywords.)
Any inquiry concerning this communication or earlier communications from the examiner should be directed to JEAN M CORRIELUS whose telephone number is (571)272-4032. The examiner can normally be reached Monday-Friday 6:30a-10p(Midflex).
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, Ann J Lo can be reached at (571)272-9767. 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.
/JEAN M CORRIELUS/Primary Examiner, Art Unit 2159 September 4, 2026
